Hi everyone !

I have a question about the RULE Hint.
One procedure does an "INSERT ... (SELECT ...)" in one table.
The select is really "heavy" (lots of tables, lots of joins).
The statistics are collected everynight one the whole schema.
The insert takes almost 2 minutes.

I put the hint RULE in the SELECT :

INSERT INTO ... (SELECT /*+ RULE */ ...)

And the insert takes just 3 seconds !
How could you explain that the rule based optimiser is more efficient than the cost based optimiser with fresh statistics ?
Thanks a lot in advance for your ideas ...